DE2353029A1 - HIGH FREQUENCY SEMICONDUCTOR AMPLIFIER EQUIPMENT AND CIRCUITS - Google Patents

Description

Patent attorney

6 Frankfurt / Main 1 Nlddastr. 52

October 22, 1973 WK / Cs

2515-RD-4768

GENERAL ELECTRIC COMPANY

1 River Road Schenectady, N.Y., U.S.A.

High frequency semiconductor amplifier devices

and circuits

The invention relates generally to three-port or two-port Semiconductor amplifier devices equipped with input possibilities and circuits for these and in particular such devices for power amplification at frequencies in the microwave range.

In the prior art, microwave power amplifiers equipped with two inputs are semiconductor devices for frequencies not known above those frequencies which are amplified by transistors, i.e. above about four gigahertz. In "Reflection type" amplifiers have become impact avalanche transit time diodes (Impact-avalanche), which are also referred to as IMPATT diodes, and electron transfer devices, which are also referred to as TED devices.

409819/07 72

Isolation of the input-side and output-side circuits or circuits obtained by circulator elements. However, the gain and stability of such reflection amplifiers using devices with two connections or with one input are very sensitive to the operating conditions. Circuit arrangements which are stable in the small signal range lead to amplifiers with a low gain in the large signal range.

In summary, the present invention provides a three-terminal or two-input semiconductor amplifier device which provides stable amplification of a high value signal at microwave frequencies, has the same simplicity as a microwave transistor in terms of circuit requirements, and still operates at frequencies, which are comparable to the operating frequencies of IMPATT diodes. These devices work as oscillators in the same range of output powers and deliver output powers with a value which is comparable to the output powers of "IMPATT" diodes, which are operated as oscillators at the same frequency.

In one embodiment of the invention, a body made of semiconductor material is provided with an emitter region of one conductivity type, a base region with the opposite conductivity type, which adjoins the emitter region and with the same forms an emitter-PN junction region, and a collector region of one conductivity type that is associated with the Base area of opposite conductivity is then arranged and forms a collector-PN transition area with the same. The collector area of the device consists of a semiconductor material in which the avalanche coefficient of the majority charge carriers is significantly greater than the avalanche coefficient of the minority charge carriers in this area. The electrodes for

409819/0772

The emitter, base and collector are each with the area for EmStiör ^ Blsls and Collector connected. The collector area owns a 'length between the collector P-N transition area and the collector electrode, which is substantially equal to half the saturation drift velocity of the majority carriers in this one Range divided by the center frequency of the facility's operating frequency band is * The collector range contains one Avalanche zone next to the collector-P-N transition area, and a drift zone which covers the remainder of the collector area includes. The length of the Lsrwinenzone is much smaller than the length of the drift zone »The average resulting concentration of activators in the collector area is designed in such a way that that when the collector is depleted of moving charge carriers by adding a voltage of suitable polarity and Amplitude between the base electrode and the collector electrode an electric field is built up in the collector area, which extends from the collector P-N junction region to the collector electrode, the electric field intensity in the Avalanche zone is largest. The resulting activator concentration in the collector area and the length of the collector area are such that the electric field intensity in the avalanche zone is less than that field intensity, which one Avalanche multiplication of the Leituftgträger in the collector area and an unlimited flow in the same would create.

A better understanding of the structure and the mode of operation of the invention can be obtained from the following description of exemplary embodiments in conjunction with the figures.

Figure! Fig. 13 is a schematic diagram of a semiconductor device that is finely connected.

FIG. 2 shows a graph of the profile or course of the resulting activator concentration of the semiconductor device according to FIG .

4098 19/07 72 '

Activator concentration on a logarithmic scale along the ordinate of the curve and the distance along the longitudinal axis of the device in a logarithmic one Scale are plotted along the abscissa of the curve.

Figure 3 shows a graph of the current flow over the applied voltage between the base and collector electrode of the device according to Figure 1 for explanation the operation of the device in order to achieve a high power gain at microwave frequencies.

FIG. 4 shows a graph of the multiplication factor for that charge which enters a hypothetical region of the profile of the electric field intensity , specifically as a function of the voltage, only the avalanche process being considered and other physical effects being neglected.

FIG. 5 shows graphs of the avalanche coefficients for electrons and holes for the material silicon and. Germanium as a function of the electric field intensity.

FIG. 6 shows a curve of the drift speed as a function the electric field intensity for electrons in semiconductor material made of silicon to explain the mode of operation the device and circuit according to Figure 1.

FIG. 7 shows graphs of the DC operating voltage and the AC voltage generated over a parallel resonance circuit including an ohmic load as a function of time and also serves to explain the mode of operation of the device and circuit according to FIG .

40981 9/0772

Figure 8 shows graphs for the current strength and the electrical charge in different parts of the device and circuit of Figure 1 depending on the Time to explain the operation of the device and circuit according to Figure 1 in conjunction with the Curves according to FIG. 7 in generating a power gain at microwave frequencies.

Figure 9 is a plan view of another embodiment of a semiconductor device according to the invention.

Figure 10 is a sectional view of the device of Figure 9 taken along section line 10-10 and showing the interior. Structure of the device according to Figure 9 and illustrates the manufacture of this facility.

FIG. 11 shows a curve of the resulting activator concentration in the device of Figure 9 as a function of the distance in the semiconductor die from the top surface to the collector electrode.

FIG. I shows a one-dimensional schematic representation of a semiconductor amplifier device 10 as an embodiment of the present invention. The device 10 contains a body of semiconductor material with a first or emitter region 11 with a very high conductivity of the N-type, a second or base region 12 with very high conductivity of the P-type, which is adjacent to the base region to form an emitter PN -Transition area 13 with the same. The device corresponds IO 'holds _a: uch a third or collector region 14 of the conductivity type N, which is then located on the base region 12 to form a collector PN junction region 15 with the same. The body of semiconductor material consists of silicon, in which the avalanche coefficient for electrons (OC _n ) is essential

409819/0772

is greater than the avalanche coefficient for holes (Ck, ) for a wide range of electrical field intensities, in that an avalanche multiplication of the line carriers occurs there. The concentration of the resulting activators or contaminants in each of the areas of the devices 10 is shown in Figure 2 as a function of the distance over the first, second and third areas. The values of the resulting activator concentration in the emitter and base regions of the device are values conventionally used for microwave transistors; that is, the emitter region has a concentration that is significantly higher than the concentration in the base region in order to provide good minority carrier injection in the base region. Of course, the base region is thin in order to obtain good transport of the injected carriers to the collector PN junction region at the high operating frequencies.

The collector region 14 contains a narrow region 16, which adjoins the collector-PN junction region 15 and is referred to as the “avalanche zone”. The collector region 14 also contains a zone 17, which is referred to as the "drift zone" and is significantly longer than the avalanche zone 16. The maximum value of the resulting activator concentration in the avalanche zone 16 is about two orders of magnitude lower than the maximum value of the resulting activator concentration in the base region 12 The peak value of the resulting activator concentration in the drift zone 17 is about two orders of magnitude lower than the resulting activator concentration in the avalanche zone 16. The contact with the drift zone 17 is established by a region 18 of a semiconductor material of the N-type with a high resulting activator concentration. This is used as a substrate and also serves as a collector electrode in order to obtain a good ohmic connection with the drift zone.

The resulting activator concentration in the region 18 can be made several orders of magnitude greater than the resulting activator concentration in the drift zone 17 and 14

have the same conductivity type. Via conductive connection provides a good connection to the area 18 with low resistance. Similar conductive terminals 20 and 21 make a good one Conductive connection with low resistance to the base region 11 or to the emitter region 12.

The resulting activator concentration along the collector region 14 is set so that when an operating voltage V is applied between the collector and base regions from a voltage source 25, the collector-PN junction region 15 is biased in the reverse direction and the collector region 14 is depleted or depleted of movable charge carriers is depleted. Under these conditions, an electric field is built up in the collector area, which has a maximum intensity at the collector-PN junction area, drops sharply in the avalanche zone and then decreases essentially linearly with the distance in the Priftzpne, since the drift zone along its length has a uniform resulting activator concentration, and drops essentially to the value zero at the interface between the drift zone ijnji «3 pm area 18. Of course, if the voltage applied in the reverse direction above the collector-PH transition area 15 is brightened beyond the value at which the collector area is currently being depleted of major carriers, the majority carriers will be depleted in the substrate 18 and an electric field will exist there. In the base area, the electric field decreases from a maximum value at the collector-PN junction area to zero. The mean value of the resulting activator concentration in the collector area and the resulting activator concentration in the avalanche area are set in relation to the resulting activator concentration in the drift zone in such a way that when the mobile carriers in the collector area become depleted, the electric field intensity in the avalanche zone causes the charge injected there with multiplied by a finite factor like this still. is explained in detail below. The length of the collector

A0 98 19/0772

area between the collector PN junction area 15 and the third oder.Kollektorelectrode 18 is set so that it substantially equal to half the saturation drift velocity is the majority carrier in that range divided by the center frequency of the facility's operating frequency band. The reasons for this can be seen from the consideration given below of the mode of operation of the device in its circuit.

A source 23 for a bias voltage V and a source 24 for RF or microwave signals to be amplified are in Series connected between the base electrode 20 and the emitter electrode. The source 23 is shown for biasing of the emitter-PN junction region 13 set up in the reverse direction, so that the device IO operates as a C-amplifier, i.e. as an amplifier in the current from the high frequency source 24 flows in the emitter-base circuit for a period of less than 180 ° of a high frequency period; this is still explained in detail in connection with FIGS. 7 and 8. Although the C mode is preferred, the device 10 is not limited to this mode of operation. A capacitor 27, an inductor 28 and a load resistor 29 are in parallel between the positive terminal of the source 23 and the collector electrode 19 switched. Functionally, this capacitor 27 also contains the capacitance of the device 10 between the collector electrode 19 and the base electrode 2O. The parallel resonance frequency of the combination is adjusted to be substantially is equal to the center frequency of the frequency band for which the device 10 is dimensioned in a particular application.

Reference is made to FIG. 3, which shows a curve 30 for direct current and. DC voltage of the reverse-biased base-collector structure of the device 10 considered alone

4098 19/07 72

shows. FIG. 3 also shows curve 31 and this represents a hypothetical direct current versus direct voltage characteristic curve for the same structure, which is to be used together with curve 37 in FIG. 4 for a better understanding of the theoretical basis of the mode of operation. The curve 37 represents the multiplication factor for that charge which enters the same base-collector structure of the device 10 over the DC voltage in a hypothetical case in which the profile of the electric field and the avalanche coefficients are independent of the magnitude of the multiplication. Points 36, 38 correspond to voltages V- and V, respectively. At the voltage V _g , the curve 37 has a vertical edge and the multiplication factor increases to infinity. The curve 31 also has a vertical gradient at the voltage V ". In a real material, for example silicon, which is used in the base-collector structure of the device 10, the collector current as shown in FIG. 3 is very low at the beginning and begins with an increase at point 32 corresponding to the voltage V- and increases further on at point 33 corresponding to voltage V. The multiplication factors at V ₁ and V are the multiplication factors attributable to the avalanche process and are essentially equal to 10 and 100. In a real material, for example silicon, however, as the multiplication factor increases, ie in

of the order of 10, the presence of the avalanche multiplication process self-generated charge that the electric field in the vicinity of the PN junction area is reduced, and at the same time causes the electric field in the Increased near the collector electrode. Such changes in the spatial field distribution at the same voltage cause that the field in the vicinity of the PN junction area drops to the field strength which is required to achieve the required To maintain multiplication factor, therefore, the same Amperage as at point 34 on hypothetical curve 31 in FIG

40 9 8 19/0772

In order to obtain FIG. 3, the actual structure requires a greater stress, as shown at point 35 in curve 30 is. Thermal effects also make a contribution to the required increase in voltage.

The curve in FIG. 3 therefore represents _{the steady-state conditions for direct current and voltage in a structure up to a voltage slightly below V 4} , which corresponds to the base-collector structure of the device 10, for example. The almost linear part of curve 30 between points V_ and V ₁ would be

2 3 2 Δ ■ 'k

de correspond to a current density of 10 to 10 A / cm. At still

4 2

higher currents, for example above 10 A / cm, the curve 30 would have a vertical slope, ie at the voltage V ₄₁ and at even higher currents it would have a negative slope and the structure would normally be irreversibly destroyed. At a bias value V ₃ corresponding to point 35 in Figure 3, the high voltage _{amplitudes can oscillate beyond the stable DC voltage limit value V 4} (but only for a short period of time) if the base and collector electrodes are connected to a suitable resonant circuit. The mean value of the current intensity is then slightly greater than the current intensity at point 35 with the same bias voltage V ₃ . This corresponds to operating the structure as an IMPATT oscillator. The multiplication factor in an IMPATT diode oscillator is greater than 10 ". For reasons which will become apparent later, the PN base collector part of the device 10 is biased with a voltage in the vicinity of V in FIGS. 3 and 4, ie with a multiplication factor in the range from about 5 to about 50. If a resonant circuit is connected to the base and collector electrodes of device 10 and the device works with the indicated collector bias, the device is stable at high frequencies, ie it will not be self-oscillating an extremely, low quiescent direct current and is able,

4098197 0 7 72

the charge appearing in the collector area in the presence of a to amplify the signal fed to the emitter electrode, as will be explained in connection with FIGS. 7 and 8.

When the collector bias voltage V is applied between the collector electrode and the base electrode of the device, the direct current flow through the device is stable, ie a given input current to the collector region generates an output current which is related to the input current via a finite multiplication factor. With such a bias, the collector region satisfies the criterion for stability, namely that the feedback integral (referred to as FBI ₀ ) is defined by the following relationship;

FBI, a 1-1 - / f ^W oC * d £ '(1)

M Jo

OC * is the effective avalanche coefficient, d *. is the thickness a differential layer and the boundaries 0 to W correspond the span of the distance over which the electric field exists. M is the multiplication factor to which reference is made in connection with FIG. Note that in this equation (1), the multiplication factor is infinite is large when the feedback integral (F.B.I.) is 1.

For an N-type semiconductor material, δC and OC. referred to as the avalanche coefficient of the electrons or the holes and the effective avalanche coefficient O & is then defined as

- ÖC βχρΛ - f (<*. "- OLJ dB 7 (2)

4098 19/0772

It should be noted that the relationship applies: CJC Zt ^ _n according to CS ^. ; the characters are accordingly in the two

P ^ *. η

To set expressions, Dph @ r for a given electric field profile the feedback integral is smaller and thus the multiplication factor M is smaller if OC <OC _n for a collector area of the N-type "compared to the case in which, for example, OL " (X _n The case in which OC ^ OC _n - therefore allows a larger voltage range from V ₁ to V - in Fig. 4 with usable ^{multiplication 1.} Such a larger, more stable DC voltage range for dl © multiplication then makes such a Semiconductor material little prone to electrical failure.

Even more important is the fact that the ratio is ° C _to

W with χ ™ ο according to the expression exp

related to the ratio of the apparent additional electrons which are located at the base-collector-PN junction area to the number of electrons actually injected. The apparent additional electrons are the result of the law of the holes, which arrive at the transition area after the holes themselves have been created shortly beforehand by the avalanche process inside the collector area. If the number of additional electrons is greater than the electrons injected by the signal, then the significance of the signal is reduced, ie the device has a large output power which is practically independent of the amplitude of the signal. In order to get a high multiplication without loss of signal quality, ie a large dynamic gain beer, eC must be much smaller

Jr

as <X _a for d @ n N-type semiliter. In summary, therefore, is di®. B & aingmg 0 ^ _n ^{>> 0} ^ for the semiconductor d © a H-Typ.s (or

409819/0772.

.- 13 -

CC »oc f _Ur the p-> type) necessary for a larger stable one

Same spinning range with moderate multiplication factor and for the larger dynamic range of the Signa! amplification.

The semiconductor silicon fulfills the requirement for an N-type material, since it has a value of oC si io <= c in the range of the electric field strengths of interest. For a P-type material, germanium with © C> oc is also suitable. the

Materials with ⁰ ^ ^ OC _for GaAs and GaP are not suitable for use in accordance with this invention. It should be noted that in many materials there is a difference between

η and OCρ is greater at lower electric field values, for example in germanium and silicon as shown in FIG Fig. 5. However, this fact cannot be used to advantage to produce great diversity by adjustment the operating voltage to a lower value, since the multiplication factor would then be too small and the facility would be unacceptable were. The silicon semiconductor material meets all requirements regarding c <- and oC and their differences

• η ρ

in the interesting multiplication area.

Reference is now made to FIG. 5, which shows the t -

en
win coefficient / for electrons and for holes in silicon ν in germanium as a function of the electrical field intensity -J ι shows. A typical value of the electric field strength in the Luwinenzone according to Figure 1 with a profile according to Figure 2 would be a current gain of about 10 / and a ratio of the avalanche coefficient of the electrons to the avalanche coefficient of the holes, of about 10 is about 3.3 χ 10 volts / cm, represented by the ordinate 39.

Reference is now made to FIG. 6 which shows the drift speed for electrons in silicon as a function of

40-98 19/0772

electric field intensity shows. With an electric field strength of about 10 volts / cm is the drift speed of

Electrons close to the saturation value of about 10 cm / sec. This value of the electric field strength is at least around one Order of magnitude smaller than the electric field strength of 3.3 10 * V / cm in the avalanche area at the bias point according to the ordinate 39 in Figure 4. Accordingly, the multiplied move Charge carriers in the avalanche zone initially with a saturation drift speed and with the reduction of The component of the field strength generated by the collector bias along the collector area can move the carrier at a speed move below the saturation drift speed. In the presence of correct high frequency operating conditions however, the charge generated by the moving avalanche increases the electric field intensity in the area through which it passes Area synchronous with the voltage change, so that for practical purposes the moving charge is the collector area traversed essentially at the saturation drift velocity. This is especially true when the density is through The charge generated by the avalanche is comparable to the resulting activator concentration and if that is necessary for the migration of the base collector space required by the charge is half the operating period of the high frequency.

After the structure of the semiconductor amplifier device 10 according to FIG. 1 and the circuit connections and likewise the movement and generation of charge carriers in the various Dividing the semiconductor device according to the bias conditions have been described, the operation of the device and circuit according to Figure 1 is explained below. The emitter bas is circuit is controlled by source 23 to operate as C amplifier biased. That is, the signal charge is from. the emitter region into the base region over a period of less than half of each operating period of high frequency is injected. Reference is now made to Figure 7, which

^ 09819/0772

shows a graph of the periodic "voltage ¥ ^ _c between the base electrode and collector electrode of the device as a function of time, the time being represented on a scale according to fractions of a high frequency period of the device. The voltage" .V ^ is the sum of the Voltage V ^ over the load impedance and the operating direct voltage represented by the ordinate "V / where the latter voltage is constant over time and corresponds to the voltage T. v'oii figure" .3 ".

At the beginning or "" at the time. Kuli ", the voltage V _{bc is} equal to the voltage V, reaches © in maximum-V. Ptstjs, at your quarter-period point and returns to the value V at ueia'HaXbperioden'-point. The voltage-V. -S then decreases. -to '© inen-' Minimum value V _K min »at the three-quarter point from un"d'ke.hrt sum

Return value V at the end of the period. "" ".

Further reference is made to the figure - ". 8, which curves for four operating parameters - in. The" semiconductor device and '. Circuit according to FIG. 1 as function: the side shows; the time is also shown here - in the form of ν-οή fractions - of a high-frequency operating period. The "" curve 41 dos Stroms ;! ,, ira. Emitter circuit ₉ the. Curve 42 for the in. Den.-Kol-l®ktos * b @ r @ I injected charge-.q., The curve 43 for di®. v © bd @ r Ltwin® generated charge q

and the curve 44 of the current i in the collector circuit are along

the ordinate plotted as a function of the tent - "The ordinates" scales for the charges are "essentially" the same and differ from the scales for the current intensities. As already indicated above, no current flows with it because of the bias of the • Eraitter-Bssis area of the device in the locked position _; The condition would be assumed if the emitter. ^ Bas-is-PK transition area is biased in the direction of passage High-frequency period. Of the "oscillation and the duration

Period ©. -.

4 0 9-8 1-9 /:. 0Π

The occurrence of the negative peak value of the high-frequency signal on the input side is set in time so that it is approximately occurs at the point in time in which the voltage V. of FIG

, dc

their negative deflection below the voltage value V begins. This is achieved by choosing the right drift path in the collector area. The curve 41 for the emitter current i _@ represents the charge which is injected in the base region. All of the charge injected at the emitter PN junction does not reach the collector PN junction.

A significant part of the injected charge goes in transit lost in the base range during high frequency operation, so that only some of the charge reaches the collector PN junction. The charge that reaches the avalanche zone is carried by the Size q. shown. The peak value for q. is a little

gen fraction of the high frequency period later than the occurrence of the peak value of the emitter circuit current i. The delay of the Peak value of q. with regard to i "represents the transit time delay

3. e '

The amplitude is reduced, and this represents a loss of charge due to recombination and diffusion. The charge q injected into the avalanche zone. is multiplied and appears as a charge which is multiplied by avalanche action in the avalanche zone. In the figure, the peak area of curve 43 is the avalanche charge q with an amplitude of about five times the amplitude of the injected charge q. The peak value of a appears after the 'peak value of the injected' charge q, au £ _o -

Sä - 3L

The avalanche charge generated in the avalanche zone essentially moves from the avalanche zone and through the drift zone with the saturation speed in the direction of the collector electrode, even during the smaller one Part of the voltage change, as shown in the

409819/0772

context rait Figure 6 was shown. When moving the Charge in the drift zone between the collector PN junction area and the collector electrode, this induces a current flow in the external circuit which is between the base electrode and the collector electrode is connected. Since the charge is a relative possesses constant amplitude and deals with a relatively constant Moving speed, it induces a relatively constant Current i between the base and collector electrodes, which

is represented by curve 44. The top formation of the Collector current i on the leading edge 45 of the same is generated

through the hole component of the current flowing in the avalanche zone has been generated and is at the collector PN junction collected. The hole component of the collector current is slightly smaller than the current drawn by the movement of electrons from the collector PN junction to the collector electrode is generated, which in turn is generated by avalanche action became. The basic component of the external current i is approximately 180 out of phase with respect to that between the base and collector electrodes appearing tension. Accordingly, power is generated in the collector area and delivered to the consumer. To one to obtain a steep rise in the collector current i and

Having a good operating efficiency is to create it necessary that the avalanche charge builds up quickly. One such Result is achieved in that an avalanche zone with a small extent in the longitudinal direction and with a such a resulting 'activator concentration is provided, which is much greater than the resulting activator concentration in the drift zone., if one made of silicon existing collector area, which also fulfills the requirement that the integral of the resulting activator concentration over the length of the area is approximately equal to

12 2 ■

2.5 x IO cm is depleted of movable load carriers then becomes at the base-collector PN junction region a sufficiently high electric field intensity 'and a low one

. 409819/0772

electric field intensity generated at the collector electrode in order to to obtain the previously described mode of operation. The integral of the resulting activator concentration in the collector area of the Devices according to FIGS. 1 and 10 and the representation according to FIGS. 2 and 11 meet the aforementioned requirements.

Under the bias V in Figure 7, the collector area is from N-type is depleted with respect to the charge carriers, and a positive space charge is built up in such a way that the electrical Field intensity at the base collector PN junction is large and is almost zero at the collector electrode. From a consideration of FIG. 7 one could apparently infer that to at a point in time of 3/4 of the period the field strength at the collector would be zero or the field strength in a part of the drift zone would be low and therefore the charge in drift zone 14 would not move at a saturation drift velocity. However, this is not the case under the correct operating conditions. Since the spatial inclination of the electric field profile is proportional to the Space charge is lifted up by the electrons generated by the avalanche process part of the positive space charge at the location of the electrons and thereby change the slope. This leads to a lowering the value of the field at the PN junction area and an increase in the value for the field at the collector connection, if the Voltage is kept constant. Therefore, the tension can be lowered to some extent without disappearing electric field anywhere inside the collector area to effect. The electrons generated by the avalanche effect therefore drift through the collector drift range at a higher rate Field strength at the rate of saturation despite the fact that during this drift time the voltage in Figure 7 experiences a downward change. The correct value for the avalanche effect generated charge is essentially equal to the resulting activator concentration in the drift zone. Determining this Correct value depends on considerations such as the intention to achieve it

9819/0772

a maximum voltage swing for the purpose of, high efficiency for the operation and the contrary intention to restrict ⁵ Lawinenmultipiikation, so that the output signal remains proportional to the input signal, as has already been discussed above in connection with the avalanche coefficient, "."'""' ■ - ■■ "" ■

Although it has been stated above that it is necessary that "the drift zone becomes impoverished in majority carriers and that electric field extends to the collector electrode, the device and the circuit arrangement can also be operated if this depletion does not extend to the collector electrode « That part of the drift zone in which there is no depletion at majority carriers, then only represents a biting impedance in the collector circuit and thus represents a loss of performance in the facility, which increases the efficiency of the facility decreased and also the preferred operating frequency the device and thus the efficiency further humiliated.

There are many loss mechanisms in the emitter area and the base area which reduce the efficiency of the injection of charge carriers into the collector area. However, these do not limit the amplification capability of the device, since according to the invention a compensating mechanism is provided, ie the avalanche multiplication of the charge which is injected into the avalanche zone. Accordingly-can the signal charge which between ■ emitter? - and. Collector area is lost for various reasons during the passage, more than is recovered through the multiplication of laundering in the avalanche zone of the collector area: -. Furthermore, the charge multiplied by the effect of La Wimen is moved ', -' in the collector area with high impedance, at high voltage, and this results in this. © ine. next © Gain in the facility / The use of a relative, alkaline drift

4 098TS / 0772

zone in the collector area makes it possible to use high voltages with a low quiescent current and thus a good degree of efficiency. Since it is desirable, a depletion of charge carriers in the collector area, with the exception of those generated by avalanche effects. To obtain the th charge, the operating mode is set so that the minimum value of the voltage V. min. does not fall below a certain small fraction of the operating voltage V drops in order to create a resulting electric field in the drift zone.

The parallel resonant circuit of FIG. 1 was represented by the lumped capacitive element 27 and the combined inductive element 28. The capacitive element 27 contains the capacitance of the device between the collector electrode and base electrode. Therefore, it enables the use of an oscillating circuit in the collector-base circle tuning out the capacity of the device and furthermore this enables an adjustment the internal impedance of the device to the load impedance 29 to which power is supplied. Although above the resonance circuit has been represented by combined elements it is easy to see that distributed elements, such as waveguides or sections of transmission lines, can be used to achieve this function.

The invention has been described above in connection with an embodiment explained, in which an avalanche zone is provided, which has a much larger resulting activator concentration in the avalanche zone than in the drift zone of the collector area. A substantially uniform resulting activator concentration in the collector area is operable provided that the above-mentioned requirements are met. The avalanche zone was shown as a narrow zone above. However, such a zone can also doze a meaningful part

4098 19/0 7 72

Make up collector area and can, for example, in the The order of magnitude of half the longitudinal extension of the collector area be.

Reference is made to Figures 9 and 10, which one Top and cross-sectional views for another semiconductor device 50 as a planar embodiment of the semiconductor device 10 according to FIG. The device 50 has areas and electrodes corresponding to the areas and electrodes the device of Figure 1. The device 50 includes a collector electrode or substrate 51 made of silicon of conductivity type N, a silicon layer 52 of conductivity type N, which is epitaxially grown on the substrate 51 and the Forms collector area 52 of the device. The device also includes a region 53 of conductivity type N, which to form the avalanche zone in the collector area from the surface the epitaxially grown layer is diffused from. the Zone between the avalanche zone 53 and the substrate 51 represents the Drift zone 60. An annular or circular area 54 of the Conductivity type P encloses the collector area and in particular the avalanche zone at a distance from the same and is formed by diffusion of P-type activators or impurities into the substrate. In operation, the ring 54 ensures that the electrical required to operate the device Field intensities are reached in avalanche zone 53, before the breakthrough values for the field intensity in the outside lying parts of the facility can be reached. A basic area 55 or a base layer 55 is epitaxially grown in the Layer by diffusion of P-type activators in this layer is formed and is in contact with the ring area the conductivity of the P-type. A variety of elongated Emitter regions 56 are in the base layer 55 of the conductivity type P formed in that silicon layers 57 of the conductive * ability type N can be grown epitaxially over the areas

409819/0772

and then from the epitaxially grown layers 47 to form the elongated or elongated emitter regions of conductivity type N are diffused Electrodes 58 of a conductor such as aluminum are attached to the N-type emitter regions and with a Emitter terminal 59 connected. Similarly, elongated electrodes 61 are made of a conductor, for example aluminum, bonded to the extended portions 62 of the base layer 55 which are in alignment with and between the extended emitter regions to form a nested one Emitter base design. The elongated electrodes 61 are connected to a base terminal 63. The emitter areas and base contact areas 62 are epitaxial with a relatively thin layer 65 of silicon dioxide on the surface of the grown layer 52 which is used in the manufacture of the device. A relatively thick layer 66 of silicon dioxide, which was used in the manufacture of the device, surrounds the emitter-base structure of the device. The emitter connection 59 and the connection to the base connection 63 are above silicon dioxide, which is on the epitaxial Layer was formed in the manufacturing process of device 50, and the emitter and base terminals are accordingly isolated from the underlying parts of the die. A suitable layer of conductor, for example Gold doped with antimony is attached to the collector electrode, i.e. the substrate 51, and serves as a collector terminal

During the manufacture of the device according to FIGS. 9 and 10 becomes a chip 51 made of silicon semiconductor material with high conductivity of the N-type and a thickness of about 0.25 mm

—2
(10 inches) received. A layer of silicon with a weak conductivity type N and a thickness of about 20 microns is epitaxially grown therein. A layer of silicon dioxide with a thickness of about 1.5 microns is thermally applied to this.

4098 19/0772

grow. A window 70 is made in the thick layer of silicon dioxide as an opening which extends to the surface of the epitaxial layer 52, and a donor impurity such as arsenic is diffused into it to form a region 53 of moderate N-type conductivity which then represents the avalanche zone of the facility. A trench or channel to the surface of the epitaxial layer is etched into the thick oxide layer 66 and completely encloses the region of the epitaxial layer 52 including the avalanche zone. The annular region 54 or guard ring is formed by diffusing an acceptor activator, for example boron, into the channel to a depth of 2 microns and then silicon dioxide is thermally grown in the channel above the guard ring to a thickness of approximately 0.5 microns. Next, in the silica including a part of the thick layer 66 is etched the oxide basis a window, which lies above the inner edge of the protective ring and is to be used to form the diffused base layer 5 ^. The base layer 55 or the region with a high conductivity type P is formed by diffusion of an acceptor activator, for example boron, into the base window. The surface of the layer in the base window is cleaned and a layer 55 of silicon dioxide is thermally grown on it. Elongated emitter windows are etched into the silicon dioxide layer to expose the surface of the layer in and on which elongated emitter regions 56 are to be formed. In the layers 57 in the emitter windows, epitaxial silicon with a high conductivity type N is applied and a diffusion is then carried out in order to form the emitter-base-PN transition regions in the base layer 55. To expose the base layer, elongated base electrode windows are etched into the silicon dioxide layer 65. The exposed parts of the base region, the epitaxially applied emitter layers and the remaining parts of the silicon dioxide layer 65 are metallized with a continuous layer of a conductor, for example aluminum. By photo-

4098 19/07 72

The lithographic mask process and etching is continuous conductive layer with a pattern for electrodes 58 connected with an emitter terminal 63, and for base electrodes 61 connected to a base terminal 63. The conductive aluminum in the case of aluminum, is sintered to allow a low resistance electrical connection to be made therewith can be. In the other opposite surface the plate or substrate 51 is alloyed with a layer of gold including an activator, for example antimony, to form the collector terminal 67 of the device.

The various processes that are used for making or forming the device used are known per se including the epitaxial growth of silicon on silicon, the thermal growth of silicon dioxide on silicon, des Forming windows in silicon dioxide layers by photolithography Masking and etching, the diffusion of activators through mask layers of silicon dioxide into an underlying layer Layer for forming areas with a geometry of the desired type and desired electrical properties, the metallization of areas to form more desirable electrical conduction paths and the like are known per se. The manufacturing steps described above can be performed under Use of such procedures can be carried out easily.

The device 50 according to Figures 9 and 10 is suitable for Use at different microwave frequencies depending on on the specific dimensions of the facility. In a structure in which the epitaxial layer has a thickness of about 20 microns and the ranges for emitter, base and collector dimensions and a profile of the resulting activator concentration, measured in a line 75 from the emitter electrode to the collector electrode as shown in FIG. 11, is suitable for operation at a frequency of about

4098 19/0 772

2 giga Hertz. To operate at higher frequencies, for example at an order of magnitude higher frequency, to allow the dimensions of the device and the other parameters of the device are then modified.

It will. Reference is made to Figure 11, which shows the profile of the resulting activator concentration of the device along a Shows line 75 from the top interface of the epitaxially grown layer down through the silicon substrate. The profile shows the resulting impurity concentration on a logarithmic scale as a function of distance d.

The emitter region 56 has a depth or extent of approximately 0.5 microns and a peak value of the resulting activator--

* 20 3
concentration of 10 atoms / cm. The base region 55, between the emitter and collector junction regions, has a depth of 0.5 microns and a peak value of the resulting activator

18 3

concentration of 10 atoms / cm. The collector region 52 is approximately 19 microns deep from the collector-PN junction region to the substrate 51 or to the collector electrode and contains the avalanche zone 53 about 2 microns wide and one Peak value of the resulting activator concentration of

16 3
10 atoms / cm, which is adjacent * to the collector-PN transition area and the drift zone 60 with a depth of about 80 microns and a relatively constant resulting activator concentration of about 5 10 atoms / cm. The high concentration of the resulting donor activators in the emitter area in relation to the concentration of the resulting acceptor activators in the base area results in a good emitter efficiency Operating voltage between base and collector electrodes and thereby

0 9 8.T9 / Ö? ? 2

the base resistance to spread is brought to a minimum and there is complete depletion in the base area Emitter-PN transition area avoided. The operation of the facility 50 is identical to the mode of operation of the device in FIG. 1 and the manner in which the device is connected in FIG a circuit is identical to the connection of the device according to Figure 1 in a circuit. The presence of the annular area or guard ring 54 made of a material with high conductivity of the P-type around the avalanche zone 53 of the Collector area 52 around, which has a relatively large radius at the cylindrical interface or the transition to the layer and a resulting donor activator concentration in of layer 52 adjacent to the interface yields that substantial is lower than in zone 53 with the conductivity * yP N, ensures that the desired electrical in the avalanche zone Field intensity is achieved before such field intensity is reached in the outlying portion 52 of the device will. The protective ring 54 expediently has essentially the same or a smaller resulting acceptor activator concentration, related to the base area.

Embodiments of the devices according to the invention have been described above described, for the silicon semiconductor material was used. It is easy to see, however, that as well well germanium semiconductor material can be used in terms of on the fact that the avalanche coefficients of holes in germanium are greater than the avalanche coefficient of electrons over a wide range of applied electric field intensity.

The device was shown above in a basic circuit. Of course, however, it can also be used in an emitter circuit. The operating mode has also been described above the device according to the invention in a circuit with loading

4098 19/0772

drove described as a C amplifier. However, the facility can Can also be used as an amplifier in A mode or B mode.

409819/0772

Claims (1)

Patent claims semiconductor device for amplifying signals in a frequency band with a body made of semiconductor material and an emitter region of one conductivity type, a base region of the opposite conductivity type, face-to-face with the emitter region, which forms an emitter-PN junction region with the same, a collector region of one conductivity type face-to-face with the area of one conductivity type, which forms a collector-PN junction area with the same, the emitter and collector junction areas being arranged at a distance from one another to produce a transistor effect between the areas and first, second and third electrodes each with the emitter-base and collector region are connected, characterized in that the collector region consists of semiconductor material in which the avalanche coefficient of the main line carriers is substantially greater than the avalanche coefficient of the minority line carriers er, wherein the collector region has a length between the collector PN junction and the third electrode which is substantially equal to half the saturation drift velocity of the majority carriers divided by the center frequency of the frequency band, and furthermore the resulting activator concentration along the collector region exhibits such a change that when the collector area is depleted on majority line carriers along its length there is an electric field in the collector area to produce a finite value of the line carrier multiplication in the area. 2. Semiconductor device according to claim 1, characterized in that the collector region contains a 4098 19/0772 avalanche zone adjacent to the collector transition region and the remaining part of the collector region represents a drift zone. 3. A semiconductor device according to claim 2, characterized in that the length of the avalanche zone is significantly smaller than the length of the drift zone. 4. Semiconductor device according to claim 2 or 3, characterized in that the mean value of the resulting concentration of the activators in the avalanche zone is substantially greater than the mean resulting concentration of the activators in the drift zone. 5. Semiconductor device according to one of claims 2 to 4, characterized in that the peak value of the resulting concentration of the activators in the avalanche zone is substantially greater than the resulting con. centering. the activators in the Dr dft zone. 6. Semiconductor device according to one of claims 2 to 5, characterized in that the resulting concentration of the activators in the drift zone is substantially uniform. . 7. Semiconductor device according to one of claims 1 to 6, characterized in that the collector region consists of silicon semiconductor material of the N conductivity type. 4098 19/077 8. Semiconductor device according to one of claims 2 to 7, characterized in that an output impedance is provided in connection with the collector region and devices for supplying a bias voltage in connection with the collector-PN junction region are provided for depletion the majority line carrier in the third area and for building up the electric field in the same and further means for supplying a signal current are provided in the circuit between the first and second electrodes for injecting a corresponding charge into the avalanche zone of the collector area, the majority line carriers being in the avalanche zone ,. which can be generated by avalanche multiplication of the injected charge move under the influence of the electric field in the drift zone to the third electrode and a current flow proportional to the signal current can be generated in the output impedance. 9. Semiconductor device according to one of claims 1-8, characterized in that the product of the ratio of the avalanche coefficients of electrons and holes in the collector region and the current multiplication in the collector region is a finite value greater than 1. Semiconductor device according to one of Claims 1 to 9, characterized in that the current multiplication is smaller than a value at which the proportional relationship of the output current to the input current in the collector region is no longer present. 409 8 19/0772 11. Semiaxter device according to one of claims 1 to 10, since the ratio of the avalanche coefficients of the majority carriers to the avalanche coefficient of the minority carriers in the avalanche zone is greater than 5. 12. Semiconductor device according to one of claims 1 to 11, since the emitter-PN transition region is equipped with such a bias that a signal current can flow for less than half a period . 13. Semiconductor device according to one of claims 1 to 112, dad u r c h g.-e-fc e η η I show η et that the output impedance is shunted to the second and third electrodes. . 1.4. * Semiconductor device according to one of claims 1 to 13, because it is clear that the output impedance contains a parallel resonant circuit. 409819/0772